OBJECT RECOGNITION METHODS IN COMPUTER VISION USING IMAGE PROCESSING LIBRARY EMGU CV

Abstract

This paper describes how object recognition is being done in delta robot station using computer vision which functionality is based on image processing library Emgu CV. In this paper there are developed some algorithms that determine rectangle form object position, color, angle of rotation and area. All these parameters allow to position and sort objects that move on conveyor belt. Each algorithm testing is being done on delta robot station that already contains video camera and artificial light source. The results show that algorithms do all necessary parameters determination and in addition these algorithms allow determining parameters for square and round objects too. Developed algorithms can be used in several disciplines, for example, in food and stationery products sorting. Application of computer vision simplifies and makes more efficient every monotone process especially when this process is being done by robots which can increase process speed and precision multiple times.